Fiveable

๐ŸงFunctional Analysis Unit 9 Review

QR code for Functional Analysis practice questions

9.3 Weak and weak* convergence

9.3 Weak and weak* convergence

Written by the Fiveable Content Team โ€ข Last updated August 2025
Written by the Fiveable Content Team โ€ข Last updated August 2025
๐ŸงFunctional Analysis
Unit & Topic Study Guides

Weak and weak* convergence are crucial concepts in functional analysis, offering alternative notions of convergence in normed spaces. They allow us to study sequences that don't converge in the usual norm topology but still exhibit useful convergence properties.

These convergence types are weaker than norm convergence but stronger than pointwise convergence. They're especially useful in infinite-dimensional spaces, where norm convergence can be too restrictive for many important sequences or series.

Weak and Weak Convergence

Weak vs weak convergence

  • Weak convergence applies to sequences in a normed space XX converging to an element in XX
    • Sequence (xn)(x_n) converges weakly to xx if for every bounded linear functional ff in the dual space Xโˆ—X^*, the sequence of scalars f(xn)f(x_n) converges to f(x)f(x)
    • Notation: xnโ‡€xx_n \rightharpoonup x
  • Weak* convergence applies to sequences in the dual space Xโˆ—X^* converging to an element in Xโˆ—X^*
    • Sequence (fn)(f_n) converges weak* to ff if for every xx in the original space XX, the sequence of scalars fn(x)f_n(x) converges to f(x)f(x)
    • Notation: fnโ‡€โˆ—ff_n \stackrel{*}{\rightharpoonup} f
    • Only applicable in the dual space, not in the original normed space
Weak vs weak* convergence, Dual space (linear algebra) - Knowino

Boundedness from weak convergence

  • Weakly convergent sequences in a normed space are always bounded
    • Consequence of the Uniform Boundedness Principle
  • If (xn)(x_n) converges weakly to xx, then there exists M>0M > 0 such that โˆฅxnโˆฅโ‰คM\|x_n\| \leq M for all nn
    • If (xn)(x_n) were unbounded, there would exist a functional ff such that โˆฃf(xn)โˆฃ|f(x_n)| is unbounded, contradicting weak convergence
  • Boundedness is a necessary condition for weak convergence, but not sufficient
    • Bounded sequences need not converge weakly (โ„“โˆž\ell^\infty with coordinate functionals)
Weak vs weak* convergence, Dual space (linear algebra) - Knowino

Weak and pointwise convergence

  • Weak* convergence of a sequence of functionals implies pointwise convergence
    • If (fn)(f_n) converges weak* to ff in Xโˆ—X^*, then for every fixed xx in XX, the sequence of scalars fn(x)f_n(x) converges to f(x)f(x)
  • Pointwise convergence does not imply weak* convergence
    • Pointwise convergence only considers convergence at each fixed point, while weak* convergence requires uniform convergence on bounded sets
  • Weak* convergence is stronger than pointwise convergence
    • Weak* convergence guarantees pointwise convergence, but not vice versa (โ„“1\ell^1 with coordinate functionals)

Examples of weak convergence types

  • Weak convergence without norm convergence:
    • Sequence (en)(e_n) in โ„“2\ell^2, where en=(0,โ€ฆ,0,1,0,โ€ฆ)e_n = (0, \ldots, 0, 1, 0, \ldots) with 11 in the nn-th position
      • Converges weakly to 00 since for any fโˆˆ(โ„“2)โˆ—f \in (\ell^2)^*, limโกnโ†’โˆžf(en)=0\lim_{n \to \infty} f(e_n) = 0
      • Does not converge in norm to 00 since โˆฅenโˆฅ=1\|e_n\| = 1 for all nn
  • Weak* convergence without norm convergence:
    • Sequence (fn)(f_n) in โ„“1\ell^1, where fn(x)=โˆ‘k=1nxkf_n(x) = \sum_{k=1}^n x_k for x=(xk)โˆˆโ„“โˆžx = (x_k) \in \ell^\infty
      • Converges weak* to f(x)=โˆ‘k=1โˆžxkf(x) = \sum_{k=1}^\infty x_k since for any xโˆˆโ„“โˆžx \in \ell^\infty, limโกnโ†’โˆžfn(x)=f(x)\lim_{n \to \infty} f_n(x) = f(x)
      • Does not converge in norm since โˆฅfnโˆฅ=n\|f_n\| = n for all nn